In order to accurately evaluate the aroma quality of watermelon grown in dry sandy soil, the volatile aroma components of 9 major watermelon cultivars planted in Jingyuan, Gansu province were identified and analyzed using headspace solid phase microextraction (HS-SPME) coupled with gas chromatography-mass spectrometry (GC-MS). The 15 major quality traits of watermelon were comprehensively evaluated by coefficient of variation and grey correlation analysis. Principal component analysis (PCA) was used to identify and analyze the major aroma components in watermelon. The grey correlation analysis showed that Kangle 6, Xisha 1 and Longke 13 ranked top three in terms of comprehensive quality. A total of 77 volatile aroma components were detected in the nine cultivars. Kangle 6 had the largest number (37) of volatile aroma components, of which 10 were unique to Kangle 6. Principal component analysis (PCA) divided the nine cultivars into four groups based on volatile aroma composition. Group Ⅰ included Longke 13, Laojincheng, Xinjincheng, Laojincheng 5, Jinhua 1, and Longxingjincheng. Group Ⅱ included Kangle 6. Group Ⅲ included Xisha 1. Group Ⅳ included Linfengjuxing. Furthermore, the major volatile aroma components of each group were identified. There was a significantly positive correlation between the number of volatile aroma components and the grey correlation coefficient of comprehensive quality, suggesting that watermelon cultivars with excellent quality traits usually contain relatively abundant volatile aroma components.